基于MMPP的802.11网络性能

提出并建立了一个基于马尔可夫调制的泊松过程MMPP(M arkovModu lated Poisson Process)和间歇泊松过程IPP(Interrupted Poisson Process)的系统模型,从新用户的到达和系统中用户发起的新业务两个方面着手,考察了在受系统门限值约束情况下的接入系统的新用户数量和系统内用户发     

港口柔性评价

港口柔性体现了港口对社会经济环境变化的反应能力和适应能力,科学的港口柔性评价是提高其柔性水平的基础.根据柔性的多维度特征和港口生产服务实际,从战略柔性、结构柔性、管理柔性、生产柔性和可持续柔性五个方面建立了港口柔性评价指标体系;选用模糊ANP方法建立评价模型,即根据各柔性指标之间的依存和反馈关系,用网络分析法(ANP)确定模糊指标权重.模糊ANP方法能够较准确地给出港口柔性的定量评价,为港口采取正确的发展策略、提高柔性水平提供参考.     

自动波竞争神经网络及其在单源最短路问题中的应用

将竞争机理引入网络的自动波产生与传播过程中，提出自动波竞争神经网络(ACNN)模型，并成功地应用于求解单源最短路问题，给出了基于ACNN的最短路求解算法. 与其他神经网络最短路算法相比，基于ACNN的最短路算法具有网络所需的神经元数目少、神经元及网络的结构简单、易于软硬件的实现、以及全并行方式计算等特点，可用于求解非对称赋权图的最短路径树问题，且其计算时间(迭代次数)仅正比于最短路径上的跃点数，与赋权图的复杂度、路径总长、边长的精度、通路总数等因素无关. 计算机仿真结果表明该算法的有效性和快速求解能力. 关键词： 单源最短路问题 自动波竞争神经网络 脉冲耦合神经网络     

在多层神经网络中用Back-Propagation算法进行方位角估测

本文介绍了多层神经网络的基本结构和主要概念,并对训练多层神经网络的Back-Propagation学习算法(即后向传递误差算法,简称后向算法)的原理和实施步骤作了详尽的分析和推导。在多层神经网络中运用这一算法,提出了平面波方位角估测的新方法。计算机模拟结果显示,这一方法是可行     

基于神经网络的三维宽场显微图像复原研究

提出一种利用BP神经网络进行三维宽场显微图像复原的非线性映射方法,将三维图像转化为二维图像进行处理,利用神经网络的学习能力,通过训练,建立含有散焦信息的二维模糊图像与二维清晰图像之间的映射关系,然后对切片堆叠进行逐幅复原,从而实现显微图像的三维复原.得到的复原图像在视觉上和定量分析上都获得了很好的效果.由于采用小规模神经网络,训练时间短,计算量小,使实时复原成为可能.     

Advanced Optical Network

The following article describes an advanced dense wavelength division multiplexing (DWDM) Optical Network developed by L-3 Photonics. The network, configured as an amplified optical bus, carries traffic simultaneously in both directions, using multiple wavelengths. As a result, data distribution is of the form peer-to-multi-peer, it is protocol independent, and it is scalable. The network leverages the rapid growth in commercial optical technologies, including wavelength division multiplexing (WDM), and when applied to military and commercial platforms such as aircraft, ships, unmanned and other vehicles, provides a cost-effective, low-weight, high-speed, and high noise-immune data distribution system.     

Self-organized Criticality in an Earthquake Model on Random Network

A simplified Olami-Feder-Christensen model on a random network has been studied. We propose a new toppling rule — when there is an unstable site toppling, the energy of the site is redistributed to its nearest neighbors randomly not averagely. The simulation results indicate that the model displays self-organized criticality when the system is conservative, and the avalanche size probability distribution of the system obeys finite size scaling. When the system is nonconservative, the model does not display scaling behavior. Simulation results of our model with different nearest neighbors q is also compared, which indicates that the spatial topology does not alter the critical behavior of the system.     

Scale-Free Properties in Traffic System

In this work, we propose a new model of evolution networks, which is based on the evolution of the traffic flow. In our method, the network growth does not take into account preferential attachment, and the attachment of new node is independent of the degree of nodes. Our aim is that employing the theory of evolution network, we give a further understanding about the dynamical evolution of the traffic flow. We investigate the probability distributions and scaling properties of the proposed model. The simulation results indicate that in the proposed model, the distribution of the output connections can be well described by scale-free distribution. Moreover, the distribution of the connections is largely related to the traffic flow states, such as the exponential distribution (i.e., the scale-free distribution) and random distribution etc.     

分立式与分布式光纤传感关键技术研究进展

光纤传感技术已广泛应用于航空航天、石油化工、电子电力、土木工程、生物医药等领域,其技术形式主要体现为分立式和分布式.分立式光纤传感技术利用光纤敏感器件作为传感器来感知被测参量的变化,光纤作为光信号的传输通道连接光纤传感器及后端的解调装置;分布式光纤传感系统基于光纤瑞利散射、拉曼散射或布里渊散射等光学效应,利用光纤本身作为传感器,可对沿途的光信号进行大范围、长距离传感.本文介绍了分立式与分布式光纤传感中主要关键技术的研究进展,并对未来的研究和发展方向进行了探讨.     

A convolutional neural network approach to calibrating the rotation axis for X‐ray computed tomography

This paper presents an algorithm to calibrate the center‐of‐rotation for X‐ray tomography by using a machine learning approach, the Convolutional Neural Network (CNN). The algorithm shows excellent accuracy from the evaluation of synthetic data with various noise ratios. It is further validated with experimental data of four different shale samples measured at the Advanced Photon Source and at the Swiss Light Source. The results are as good as those determined by visual inspection and show better robustness than conventional methods. CNN has also great potential for reducing or removing other artifacts caused by instrument instability, detector non‐linearity, etc. An open‐source toolbox, which integrates the CNN methods described in this paper, is freely available through GitHub at tomography/xlearn and can be easily integrated into existing computational pipelines available at various synchrotron facilities. Source code, documentation and information on how to contribute are also provided.